A Genetic Algorithm Approach to Design Principles for Organic Photovoltaic Materials

The increase in the efficiency of organic photovoltaic (OPV) devices relies on understanding the underlying science of several interconnected physical mechanisms that prevent the success of 1D optimization strategies. Here, a combination of kinetic Monte Carlo simulations of exciton dynamics with a genetic algorithm to automatically optimize the external quantum efficiency of donor–acceptor interfaces under different scenarios is employed. Simulations include phenomena from light absorption to exciton diffusion, dissociation, radiative recombination, and internal conversion, thus modeling the main physical processes that define the overall efficiency of an OPV up to charge separation. It is shown that when internal conversion is kept in check, the combination of optimal transition dipole moments and absorption energies points at low bandgap polymers as the most promising candidates for donor materials. However, when non-radiative deexcitation mechanisms are stronger, the optimization strategy shifts toward higher bandgaps, focusing rather on increasing the fluorescence quantum yield of the donor. Finally, the approach shows that adjusting the energy levels of the acceptor so that exciton transfers across the interface become negligible produces important gains in efficiency and at the same time reduces the system's dependence on large electronic couplings. The findings indicate pathways for engineering highly efficient organic interfaces.